Subsetting dates in R - r

Question

I have the following data
dat <- structure(list(Datetime = structure(c(1261987200, 1261987500,
1261987800, 1261988100, 1261988400), class = c("POSIXct", "POSIXt"
), tzone = ""), Rain = c(0, -999, -999, -999, -999)), row.names = c(NA,
5L), class = "data.frame")
The first column contains the dates (year, month, day, hour). The second column is Rainfall.
The dates are not continuous. Some of the dates with missing Rainfall were already removed.
I would like to ask what is the best way of subsetting this data in terms of Year, Day, month or hour?
For example, I just want to get all data for July (month = 7). What I do is something like this:
dat$month<-substr(dat$Datetime,6,7)
july<-dat[which(dat$month == 7),]
or if its a year, say 2010:
dat$year<-substr(dat$Datetime,1,4)
dat<-which(dat$year == 2010),]
Then convert them into numeric types.
Is there an easier way to do this in R? the dates are already formatted using POSIXlt.
I'll appreciate any help on this.
Lyndz

Answer 1

If you want to convert the Datetime to year or month (numeric), you can try format like below
df1 <- transform(
df,
year = as.numeric(format(Datetime,"%Y")),
month = as.numeric(format(Datetime,"%m"))
)
which gives
Datetime Rain year month
1 2009-12-28 09:00:00 0 2009 12
2 2009-12-28 09:05:00 -999 2009 12
3 2009-12-28 09:10:00 -999 2009 12
4 2009-12-28 09:15:00 -999 2009 12
5 2009-12-28 09:20:00 -999 2009 12
If you want to subset df1 further by year (for example, year == 2010), then
subset(
df1,
year == 2010
)

Answer 2

You can use the lubridate package and associated month and year functions.
library(tidyverse)
library(lubridate)
df <- structure(list(
Datetime = structure(
c(1261987200, 1261987500,
1261987800, 1261988100, 1261988400),
class = c("POSIXct", "POSIXt"),
tzone = ""
),
Rain = c(0,-999,-999,-999,-999)
),
row.names = c(NA,
5L),
class = "data.frame") %>%
as_tibble()
df %>%
mutate(month = lubridate::month(Datetime),
year = lubridate::year(Datetime))
Output:
# A tibble: 5 x 4
Datetime Rain month year
<dttm> <dbl> <dbl> <dbl>
1 2009-12-28 16:00:00 0 12 2009
2 2009-12-28 16:05:00 -999 12 2009
3 2009-12-28 16:10:00 -999 12 2009
4 2009-12-28 16:15:00 -999 12 2009
5 2009-12-28 16:20:00 -999 12 2009